Overexpression of the SARS-CoV-2 receptor ACE2 is induced by cigarette smoke in bronchial and alveolar epithelia

Abstract

Angiotensin-converting enzyme 2 (ACE2) has been identified as the functional receptor of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and a target for disease prevention. However, the relationship between ACE2 expression and its clinical implications in SARS-CoV-2 pathogenesis remains unknown. Here, we explored the location and expression of ACE2, and its correlation with gender, age, and cigarette smoke (CS), in a CS-exposed mouse model and 224 non-malignant lung tissues (125 non-smokers, 81 current smokers, and 18 ex-smokers) by immunohistochemistry. Moreover, the correlations of ACE2 with CS-induced oxidative stress-related markers, hypoxia-inducible factor-1α (HIF-1α), inducible nitric oxide synthase (iNOS), and 4-hydroxynonenal (4-HNE) were investigated. Chromatin immunoprecipitation and luciferase reporter assays identified the cause of ACE2 overexpression in human primary lung epithelial cells. We demonstrated that ACE2 was predominantly overexpressed on the apical surface of bronchial epithelium, while reduced in alveolar epithelium, owing to the dramatically decreased abundance of alveolar type II pneumocytes in CS-exposed mouse lungs. Consistent with this, ACE2 was primarily significantly overexpressed in human bronchial and alveolar epithelial cells in smokers regardless of age or gender. Decreased ACE2 expression was observed in bronchial epithelial cells from ex-smokers compared with current smokers, especially in those who had ceased smoking for more than 10 years. Moreover, ACE2 expression was positively correlated with the levels of HIF-1α, iNOS, and 4-HNE in both mouse and human bronchioles. The results were further validated using a publicly available dataset from The Cancer Genome Atlas (TCGA) and our previous integrated data from Affymetrix U133 Plus 2.0 microarray (AE-meta). Finally, our results showed that HIF-1α transcriptionally upregulates ACE2 expression. Our results indicate that smoking-induced ACE2 overexpression in the apical surface of bronchial epithelial cells provides a route by which SARS-CoV-2 enters host cells, which supports clinical relevance in attenuating the potential transmission risk of COVID-19 in smoking populations by smoking cessation. © 2020 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Keywords: ACE2; COVID-19; SARS-CoV-2; alveolar epithelial cells; bronchial epithelial cells; cigarette smoke.

Figure 1

Overall expression of ACE2 is downregulated in lung alveolar epithelial cells of mice exposed to CS. (A) Representative images of ACE2, SFTPC, and CD68 immunostaining of lung alveolar sections from mice in non‐smoking (1 month, n = 8; 4 months, n = 8; 8 months, n = 8) and cigarette smoke‐exposure (CS, 1 month, n = 8; 4 months, n = 7; 8 months, n = 5) groups. (B) Staining index of ACE2 in lung alveolar epithelial cells from different exposure groups. (C) The percentage of ATII cells as assessed by SFTPC staining of alveolar epithelial cells. Exact P values are shown above each graph. n.s., not significant. Scale bars: upper panel, 100 μm; lower panel, 20 μm.

Figure 2

ACE2 is significantly overexpressed in bronchial epithelial cells of mice exposed to CS. (A) Representative images of ACE2, CC10, and Ac‐α‐tubulin immunostaining of lung bronchial tissues from mice in non‐smoking (1 month, n = 8; 4 months, n = 8; 8 months, n = 8) and cigarette smoke‐exposure (CS, 1 month, n = 8; 4 months, n = 7; 8 months, n = 5) groups. (B) Staining index of ACE2 in bronchioles from the different groups. Exact P values are shown above each graph. Scale bars: upper panel, 100 μm; lower panel, 20 μm.

Figure 3

ACE2 expression positively correlates with the levels of CS‐induced oxidative stress‐related markers HIF‐1α, iNOS, and 4‐HNE in mouse bronchioles. (A) Representative images of ACE2, HIF‐1α, iNOS, and 4‐HNE immunostaining in lung bronchial tissues in non‐smoking (n = 8) and CS (n = 5) mouse groups exposed for 8 months. (B–D) Levels of HIF‐1α (B), iNOS (C), and 4‐HNE (D) in ACE2‐low (n = 29) and ACE2‐high (n = 15) tissues. (E–G) Correlation of ACE2 expression with the levels of HIF‐1α (E), iNOS (F), and 4‐HNE (G) in mouse bronchioles. Exact P values are shown above each graph. n.s., not significant. Scale bars: upper panel, 100 μm; lower panel, 20 μm.

Figure 4

ACE2 is significantly overexpressed in human bronchial epithelial cells of non‐malignant lung tissues with a smoking history. (A) Representative images of ACE2 immunostaining of bronchi of lung tissues in non‐smoker (n = 125), current smoker (pack‐years < 20, n = 43; pack‐years ≥ 20, n = 38), and ex‐smoker (smoking cessation < 10 years, n = 10; smoking cessation ≥ 10 years, n = 8) groups classified by smoking pack‐years and cessation time. (B) Statistical analysis of ACE2 staining index classified by different smoking status. (C) mRNA expression of ACE2 in non‐smoker (n = 7), current smoker (n = 35), and ex‐smoker (ceased smoking less than 10 years ago, n = 17; ceased more than 10 years ago, n = 26) groups of normal lung tissues from the TCGA dataset. (D) mRNA expression of ACE2 in non‐smoker (n = 66) and current smoker (n = 8) groups of normal lung tissues from the AE‐meta dataset. Exact P values are shown above each graph. n.s., not significant. Scale bars: upper panel, 200 μm; lower panel, 50 μm.

Figure 5

ACE2 expression positively correlates with the levels of CS‐induced oxidative stress‐related markers HIF‐1α, iNOS, and 4‐HNE in human bronchial epithelial cells of lung tissues. (A) Representative images of ACE2, HIF‐1α, iNOS, and 4‐HNE immunostaining of human lung bronchial tissues in non‐smoker (n = 125) and smoker (n = 99) groups are shown. (B–D) Levels of HIF‐1α (B), iNOS (C), and 4‐HNE (D) in non‐smoker (n = 125) and smoker (n = 99) tissues. (E–G) Levels of HIF‐1α (E), iNOS (F), and 4‐HNE (G) in ACE2‐low (n = 130) and ACE2‐high expression (n = 94) tissues. (H–J) Correlation of ACE2 expression with the levels of HIF‐1α (H), iNOS (I), and 4‐HNE (J) in lung bronchial tissues with smoking history. Exact P values are shown above each graph. Scale bars: upper panel, 200 μm; lower panel, 50 μm.

Figure 6

HIF‐1α contributes to ACE2 expression in primary human lung epithelial cells. (A–C) Western blotting for ACE2 expression after treatment of PHSAE and PHATII cells with (A) CoCl₂, (B) cytomix, and (C) 4‐HNE. (D) Western blotting for ACE2 in PHSAE and PHATII cells with different expression of HIF‐1α. (E) HIF‐1α‐binding motifs in the putative promoter region of ACE2 in the UCSC genome browser. (F) ChIP analysis of the binding sites of HIF‐1α in the promoter region of ACE2 in PHSAE and PHATII cells. IgG was used as a negative control. Each bar represents the mean ± SD of three independent experiments. *p < 0.05. (G) Luciferase reporter activity of the ACE2 promoter [full length (FL) or truncations] in HIF‐1α‐overexpressing lung cells. Each bar represents the mean ± SD of three independent experiments. *p < 0.05.